POLICY WIRE FACT CHECK: Sam Altman Claims 38,000 ChatGPT Queries Use as Much Water as Growing a Single Almond
The Claim A viral claim circulating on social media and news platforms asserts that Sam Altman, the former CEO of OpenAI, stated that 38,000 ChatGPT queries use as much water as growing a single...
The Claim
A viral claim circulating on social media and news platforms asserts that Sam Altman, the former CEO of OpenAI, stated that 38,000 ChatGPT queries use as much water as growing a single almond. The claim gained traction through various online outlets, including Yahoo Tech, which published a fact-check article titled Fact check: Sam Altman says 38,000 ChatGPT queries use as much water as growing a single almond. The specific statement attributed to Altman was widely shared, often without context or verification, sparking public concern about the environmental impact of AI technologies.
The claim originated from an article linked via Google News Fact Check, which cited the statement as part of a broader discussion on the energy and resource consumption of large language models. The viral nature of the claim was amplified by its alarming implication—that the operation of AI systems could have significant environmental consequences, particularly in water-scarce regions. The original article did not provide direct video or audio evidence of Altman making the statement, but it referenced a potential quote that was later scrutinized for accuracy and intent.
The Details & Investigation
Upon investigation, no verifiable record of Sam Altman making the specific claim about 38,000 ChatGPT queries consuming as much water as a single almond was found in any official transcript, interview, or public speech. Multiple reputable sources, including OpenAI’s official website, press releases, and interviews with Altman, do not contain this exact statement. Additionally, no video or audio recording of Altman stating this figure has been identified in public archives or verified by credible news organizations.
The claim appears to be based on a misattribution or misinterpretation of a broader conversation about the environmental impact of AI. In recent years, there have been discussions around the carbon footprint and energy use of large-scale AI models, but these conversations typically focus on electricity consumption rather than water usage. While some studies have explored the indirect water consumption associated with data centers (due to cooling systems), there is no established scientific consensus or authoritative report that links 38,000 ChatGPT queries to the equivalent of one almond’s water requirement.
📄 POLICY WIRE WHITEPAPER PUBLISHED: PAKISTAN’S NATIONAL SECURITY POLICY PRIORITIES
Furthermore, the claim may have been influenced by a combination of technical inaccuracies and sensationalized reporting. The phrase “growing a single almond” is often used in environmental contexts to illustrate the water-intensive nature of certain agricultural practices, but applying it to AI operations lacks empirical support. Given the absence of a direct quote from Altman and the lack of credible data backing the specific figure, the claim likely falls under the category of MISINFORMATION—unintentional error or misleading context—rather than deliberate disinformation.
Investigative journalists and fact-checkers at PolicyWire have cross-referenced multiple sources, including academic research on AI energy use, industry reports from OpenAI and other tech firms, and statements from environmental experts. None of these sources corroborate the specific claim attributed to Altman. Instead, they emphasize the need for more nuanced understanding of AI’s environmental impact, which involves complex factors beyond simple water usage metrics.
The Verdict
The claim that Sam Altman stated 38,000 ChatGPT queries use as much water as growing a single almond is FALSE. No credible source confirms that Altman made this statement, and there is no verified data supporting the specific water consumption figure attributed to AI queries. The claim appears to stem from a misattribution or misunderstanding of broader discussions about the environmental impact of AI technologies. While concerns about the sustainability of large-scale computing are valid, the specific figure cited in the viral post is not supported by factual evidence.
Given the lack of direct evidence and the absence of a verified source, the claim is best classified as MISINFORMATION—likely the result of an error in reporting or an overgeneralization of a more complex issue. It does not meet the threshold of deliberate deception required to be labeled DISINFORMATION. However, the spread of such claims highlights the importance of critical evaluation of information, especially when it relates to high-profile figures and emerging technologies like AI.
Counter-misinformation & disinformation investigation conducted by PolicyWire Editorial Desk (PW).
